dhimmel/pubmedpy

Utilities for interacting with NCBI EUtilities relating to PubMed

Utilities for interacting with NCBI EUtilities relating to PubMed

pubmedpy is a Python 3 package for interacting with PubMed's E-Utilities API as well the PubMed Central OAI-PMH Service

Projects

This package was created while working on https://github.com/greenelab/iscb-diversity. However, it draws on code and experience from several pre-existing projects. Parts of the codebase derives from or supported the analyses for the following analyses:

1. Analysis of ISCB honorees and keynotes reveals disparities
   Trang T. Le, Daniel S. Himmelstein, Ariel A. Hippen Anderson, Matthew R. Gazzara, Casey S. Greene
   Manubot (2020-01-30) https://greenelab.github.io/iscb-diversity-manuscript/

2. On author versus numeric citation styles
   Daniel Himmelstein
   Satoshi Village (2018-03-12) https://blog.dhimmel.com/citation-styles/

3. The history of publishing delays
   Daniel Himmelstein
   Satoshi Village (2016-02-10) https://blog.dhimmel.com/history-of-delays/

4. Publication delays at PLOS and 3,475 other journals
   Daniel Himmelstein
   Satoshi Village (2015-06-29) https://blog.dhimmel.com/plos-and-publishing-delays/

pubmedpy's Issues

Empty affiliation strings from PMCID

I'm trying to extract affiliation information from PMCID. For example, for PMC6986235, I tried the following:

from lxml.etree import tostring

art = get_frontmatter_etree_via_api('PMC6986235')
print(tostring(art, encoding = 'unicode'))

Part of the output contains the affiliation of the corresponding author:

<aff id="A1">Georgetown University, Department of Oncology and Lombardi
Comprehensive Cancer Center, Washington, DC, 20007.</aff>

However, when I tried

extract_authors_from_article(art)

all affiliations is empty:

[{'pmcid': 'PMC6986235',
  'position': 1,
  'fore_name': 'Ziling',
  'last_name': 'Fan',
  'corresponding': 0,
  'reverse_position': 3,
  'affiliations': []},
 {'pmcid': 'PMC6986235',
  'position': 2,
  'fore_name': 'Yuan',
  'last_name': 'Zhou',
  'corresponding': 0,
  'reverse_position': 2,
  'affiliations': []},
 {'pmcid': 'PMC6986235',
  'position': 3,
  'fore_name': 'Habtom W.',
  'last_name': 'Ressom',
  'corresponding': 0,
  'reverse_position': 1,
  'affiliations': []}]

It is possible that we can't extract this information because of the way journals deposited the metadata. I just wanted to make sure that there is not a better alternative than skipping these articles entirely.

Retrieve number of citations given PMIDs

It would be very useful to have a function that takes PMIDs and returns their corresponding number of citations. Perhaps interacting with the PubMed API following this guideline?
https://www.ncbi.nlm.nih.gov/pmc/tools/cites-citedby/